Disk consolidation failed

[trackstar]

Hello,

For the second straight day, the disk consolidation has failed (stuck snapshot). We had successful with snapshot hunter until the last few days. Prior to this version we had to create a manual snapshot and delete all the snapshot to force the disk consolidation (per VMware suggestions). The process for disk consolidation would take a long time. I am assuming we might have to stop the backup (don't run any backup while we perform manual disk consolidation) and do what VMware suggested before.


Version 8.0.0.917


Thanks,
TT

[foggy]

Yes, you can try that. What kind of snapshot it is (any chance it is a temp snapshot created during the backup?)? I'd also recommend contacting support for a closer look.

[dellock6]

By the way, build 917 is the plain v8 release, even if maybe not related to your issue, you may want to upgrade to the latest Update 2 version.

[trackstar]

Foggy,

Looks like a snapshot from backup cause when we do the manual consolidation, it clears the alert. When backup is run & finished that is when the consolidation message pops up. This is a 2010 Exchange machine.

[foggy]

Temporary snapshots created during backup should be removed during the next job run (if not removed after the backup is over). Please investigate closer with support.

[trackstar]

This morning, the Veeam back job finished and it triggered the disk consolidation. The process is still going since ~ 8 am. It is at 87% now. If the disk consolidation doesn't finished by the time the back job starts again, I should unscheduled the job right?

Thanks again.

[resruss]

Is the host running ESXi 5.5 by any chance?
After a similar situation to you (every couple of days needed consolidating), we ended up finding that VMware changed the way they remove snapshots from 5.1 to 5.5
It checks to see how long it will stun the VM for during snapshot removal and if it's longer than that value (12 seconds by default from memory), it aborts.

Some reading that explains it in more depth:
http://kb.vmware.com/selfservice/micros ... Id=2082886
https://communities.vmware.com/thread/493639

Edit: The default value for maxConsolidateTime is 6, but it's a coefficient of 12, we ended up bumping the value to 30, which will allow the stun to occur for 60 seconds before bailing.

We still get the conslidation needed every now and again, but it corrects itself the next time a backup runs.